AI emotion recognition does not work. Period.
The tech isn’t ready. The research isn’t there. Privacy and human rights are at risk.
More research is good, but Apple needs to give users freedom to opt in/out of emotion recognition, and more transparency and accountability.
The tech isn’t ready. The research isn’t there. Privacy and human rights are at risk.
More research is good, but Apple needs to give users freedom to opt in/out of emotion recognition, and more transparency and accountability.
https://twitter.com/rolfewinkler/status/1440319837318746116
What are the harms?
Imagine exam tech that fails you if it thinks you’re too nervous. Or security surveillance that flags you for screening if you seem agitated. Or chat AI that notifies police if you seem to be in a mental health crisis.
These all exist! And none of them work!
Imagine exam tech that fails you if it thinks you’re too nervous. Or security surveillance that flags you for screening if you seem agitated. Or chat AI that notifies police if you seem to be in a mental health crisis.
These all exist! And none of them work!
• • •
Missing some Tweet in this thread? You can try to
force a refresh
